We’ve never looked favorably on red-light cameras. National studies questioned their ability to prevent crashes. Once installed, the systems often appeared to be little more than cash machines for revenue-starved governments.

Still, one nearby county’s recent experience is noteworthy: It suggests the devices actually do stop people from speeding through intersections and putting other drivers at risk. Behavior may change for the better.

Manatee County last week reported a significant drop in red-light violations at intersections with cameras.

According to the Bradenton Herald, the county’s four red-light camera intersections recorded 367 violations in May, an increase from the 255 in April. However, the numbers were way down from the 661 reported in October 2012, the first full month after cameras were installed. Even more impressive, violations had dropped from a high of 1,297 in November.

It’s hard to escape the impression the program is making critical intersections safer. Although the sample size is small, habits do seem to be changing. The sampling only involved violations — not crashes — but it does add weight to the pro-camera side of the debate.

Florida’s evidence does seem to be moving in that direction.

Late last year, the state Department of Highway Safety and Motor Vehicles released an analysis of data from 73 police agencies with red-light cameras from July 1, 2011, to June 30, 2012. Of the total, 56 percent saw an overall decrease in crashes. Fifteen percent reported an increase and 10 percent saw no change. (Nineteen percent didn’t report).

Some 44 percent saw a decrease in side-impact crashes, the most dangerous type of crash. Eleven percent saw an increase, 15 percent saw no change.

As for rear-end collisions, 41 percent saw a decrease, 22 percent an increase and 7 percent no change. This contradicts previous national studies indicating red-light cameras deterred T-bone crashes but increased rear-end crashes. (Fearing a ticket, people stomp on their brakes and get hit from behind.)

Manatee County’s experience also counters the cash-machine claim. In sum, the county netted $396, a minimal amount. Then again, the Tampa Bay Times reported last year that Tampa had collected a whopping $2.3 million from its cameras — above the $1 million it paid a vendor to operate the system. The state highway department said nearly 1 million people received tickets over the study period. The fine for a non-moving violation is $158.

A bill recently signed by Gov. Rick Scott addresses a couple of small issues with red-light cameras. As of July 1, motorists cannot be ticketed for taking a right on red, if they come to a full stop first. It also extends the time motorists have to respond to the tickets that come by mail. And it requires cities and towns to establish an appeal process.

Overall, we’re still not convinced. But Manatee’s statistics help tip the scales toward the public safety argument.